Download Population density of North American elk

Oecologia (2009) 161:303–312
DOI 10.1007/s00442-009-1376-z
PLANT-ANIMAL INTERACTIONS - ORIGINAL PAPER
Population density of North American elk: effects on plant
diversity
Kelley M. Stewart Æ R. Terry Bowyer Æ
John G. Kie Æ Brian L. Dick Æ Roger W. Ruess
Received: 6 November 2008 / Accepted: 4 May 2009 / Published online: 30 May 2009
Ó Springer-Verlag 2009
Abstract Large, herbivorous mammals have profound
effects on ecosystem structure and function and often act as
keystone species in ecosystems they inhabit. Densitydependent processes associated with population structure
of large mammals may interact with ecosystem functioning
to increase or decrease biodiversity, depending on the
relationship of herbivore populations relative to the carrying capacity (K) of the ecosystem. We tested for indirect
effects of population density of large herbivores on plant
species richness and diversity in a montane ecosystem,
where increased net aboveground primary productivity
(NAPP) in response to low levels of herbivory has been
reported. We documented a positive, linear relationship
between plant-species diversity and richness with NAPP.
Structural equation modeling revealed significant indirect
relationships between population density of herbivores,
NAPP, and species diversity. We observed an indirect
effect of density-dependent processes in large, herbivorous
mammals and species diversity of plants through changes
in NAPP in this montane ecosystem. Changes in species
diversity of plants in response to herbivory may be more
indirect in ecosystems with long histories of herbivory.
Those subtle or indirect effects of herbivory may have
strong effects on ecosystem functioning, but may be
overlooked in plant communities that are relatively resilient to herbivory.
Keywords North American elk Herbivore optimization NAPP Species diversity Species richness Structural equation models
Introduction
Communicated by Janne Sundell.
K. M. Stewart (&)
Natural Resources and Environmental Science,
University of Nevada Reno, 1000 Valley Rd/MS 186,
Reno, NV 89512, USA
e-mail: [email protected]
R. T. Bowyer J. G. Kie
Department of Biological Sciences, Idaho State University,
921 South 8th Avenue, Stop 8007, Pocatello, ID 83209, USA
B. L. Dick
United States Forest Service, Pacific Northwest Research
Station, 1401 Gekeler Lane, La Grande, OR 97850, USA
R. W. Ruess
Institute of Arctic Biology and Department of Biology and
Wildlife, University of Alaska Fairbanks, Fairbanks, AK 99775,
USA
Large herbivores often act as keystone species (sensu Molvar
et al. 1993; Simberloff 1998) in ecosystems they inhabit, and
may either increase or decrease plant diversity, depending on
their population density relative to the carrying capacity (K)
of the ecosystem. Large mammals have unique population
dynamics, they exhibit strong density-dependent population
growth (McCullough 1979; Stewart et al. 2005), and unique
life-history characteristics; those species are not simply
small mammals scaled large (Caughley and Krebs 1983).
Dynamics and density-dependent processes associated with
populations of large mammals may have cascading effects
on ecosystem function and community structure, including
changing levels of biodiversity (Person et al. 2001; see
Bowyer et al. 2005 for review). DeCalesta and Stout (1997)
reported that interactions between populations of large
mammals and their environments should be standardized to
K to provide understanding of effects of large mammals on
123
304
their environments. High-levels of herbivory, from populations of large herbivores at or near K, often lead to declines in
plant species diversity and loss of highly palatable species
from the plant community (Olff and Ritchie 1998; Rooney
2001; Vellend 2004; McShea 2005; Nicholson et al. 2006).
Indeed, loss of plant diversity may have cascading effects on
other trophic levels in the ecosystem (Berger et al. 2001;
Terborgh et al. 2001). Conversely, herbivory by large
mammals at low to moderate densities has led to positive
influences on ecosystem functioning, such as increasing net
aboveground primary productivity (NAPP) and increased
nutrient cycling with inputs of urine and feces—a process
known as herbivore optimization (McNaughton 1979; Ruess
and McNaughton 1987; Ruess et al. 1989; Hik and Jefferies
1990; Frank and McNaughton 1993; Stewart et al. 2006).
Thus, plant responses to herbivory depend on the intensity of
foraging, which is determined by population density of
herbivores relative to carrying capacity (K) of the ecosystem
(Kie et al. 2003; Persson et al. 2005; Stewart et al. 2006). We
define low to moderate levels of herbivory as population
density of herbivores that is at or below the maximum
number of recruits (e.g., maximum sustained yield—MSY)
for a population with density-dependent growth (McCullough 1979; Fowler 1981).
Species composition is likely a major determinant of
stability, primary productivity, nutrient dynamics, invasibility, and other critical attributes of ecosystems (Tilman
1999). Low levels of herbivory may upset competitive
interactions among plants and prevent particular species
from dominating a community (Pastor and Cohen 1997;
Jacobs and Naiman 2008). Thus, at low levels, grazing or
browsing reduces biomass and canopy cover, primarily of
competitively dominant plants, which leads to increases in
spatial heterogeneity, and in turn allows more species of
plants to coexist (Olff and Ritchie 1998; Jacobs and Naiman
2008). Conversely, high levels of herbivory often led to
declines in plant species richness and diversity, as plant
species are reduced or eliminated by herbivory or by trampling, and generally those species that are resistant to herbivory or trampling are more likely to remain in the
ecosystem (Olff and Ritchie 1998; Vellend 2004; McShea
2005; Nicholson et al. 2006; Jacobs and Naiman 2008).
Finally, communities that are more diverse often are more
productive and resistant to disturbance, including effects
from herbivory (Tilman 1999; Collins et al. 1998). Vegetation plots with high species diversity in Serengeti grasslands
were more resistant to grazing and had higher productivity
than less-diverse plots (McNaughton 1983, 1985).
North American elk (Cervus elaphus) are an especially
good model for examining effects of herbivores at high and
low density in montane ecosystems. Life-history characteristics of elk are consistent with species that exhibit strong
density dependence; elk populations show reductions in
123
Oecologia (2009) 161:303–312
physical condition and pregnancy rates at high population
density (Stewart et al. 2005). These large mammals occur
extensively across the intermountain west of North America
and consume a variety of forages, thereby affecting species
richness and diversity of most functional groups of plants
(Kie et al. 2003; Stewart et al. 2003). Stewart et al. (2006),
in a companion study to this paper, reported herbivore
optimization by North American elk in this same montane
ecosystem, and observed increases in NAPP at low levels of
herbivory in forbs, graminoids, and shrubs. Herbivory by
elk has also been implicated both in increasing and
decreasing rates of nutrient cycling in soils (Singer and
Schoenecker 2003; Schoenecker et al. 2004).
Although Stewart et al. (2006) documented herbivore
optimization, those authors observed no direct link between
levels of herbivory and species diversity of the plant community. Our objective was to further investigate the relationship between population densities of North American
elk with changes in the plant community, and to examine
indirect effects of herbivory by elk on species composition
of plants. The relationship between herbivore population
dynamics and species composition is potentially more
subtle than directly changing the species composition of the
ecosystem, particularly in ecosystems with a long history of
herbivory by native and domestic ungulates. Thus, herbivory may change species diversity of plants indirectly by
changing the productivity of those plants. We hypothesize
that those indirect effects of herbivores on species composition at low population densities of herbivores modify plant
communities more strongly than selective removal of palatable species. Furthermore, we hypothesize that the relationship between species composition of the plant
community and herbivory is mediated through changes in
NAPP with varying levels of herbivory. Based on those
hypotheses, we predict that, in areas of low herbivory by elk
where NAPP has been documented (Stewart et al. 2006) to
be increased compared with areas of no or high herbivory
(e.g. the hump of the herbivore optimization curve), we will
observe increased species richness and diversity of plants.
Materials and methods
Study area
We conducted research from 1999 through 2001 on the
Starkey Experimental Forest and Range (hereafter Starkey)
of the U.S. Forest Service. Starkey (45°130 N, 118°310 W) is
situated in the Blue Mountains of northeastern Oregon,
USA, and is located 35 km southwest of La Grande,
Oregon. Elevations on Starkey range from 1,120 to
1,500 m. Starkey encompasses 10,125 ha, and since 1987
has been surrounded by a 2.4-m fence that prevents
Oecologia (2009) 161:303–312
immigration or emigration of large herbivores, including
migration to traditional winter ranges (Rowland et al.
1997; Stewart et al. 2002, 2006).
We restricted our experiment to the northeast area on
Starkey (Fig. 1), which encompassed 1,452 ha, and was
separated from the remainder of the study area by the same
high fence (Stewart et al. 2002). The northeast area was
divided into 2 study sites with the 2.4-m fence, east (842 ha)
and west (610 ha), to accommodate experimental comparisons of two population densities of elk (Fig. 1). We divided
the northeast area in a manner that resulted in plant communities being equal in proportions in eastern and western
areas (Stewart et al. 2002). Such study sites are sufficiently
large to allow natural movements and other behaviors of
large herbivores (Hirth 1977; McCullough 1979; Stewart
et al. 2002). Stewart et al. (2002) examined locations of elk
in the northeast study area and reported no significant
effects of the high fence on habitat selection by elk.
Elk no longer migrate off the study area to traditional
winter ranges because of the fence; consequently, animals
were maintained throughout winter in a holding area and
were fed only a maintenance diet of alfalfa hay (Rowland
et al. 1997; Stewart et al. 2005, 2006). Elk were held on the
winter feedground from early December until late April
(Rowland et al. 1997; Stewart et al. 2006). Few elk
remained on the study area during winter; thus, herbivory
305
by elk in our study was restricted to spring, summer, and
autumn (Stewart et al. 2002, 2005, 2006). Because of the
design of the Starkey Project, the elk population can be
manipulated nonlethally by releasing specific numbers of
animals into each study area via a system of fenced
alleyways between the feedground and each of the study
areas on Starkey (Main Study, Campbell, Northeast east,
Northeast west; Stewart et al. 2005, 2006).
Data were collected on physical condition, body mass,
and pregnancy rates of female elk as they entered the winter
feedground each year (Rowland et al. 1997; Stewart et al.
2002, 2005). Summer is the time of resource acquisition for
ungulates in seasonal environments (Mautz 1978), and elk
in this study were documented to show density-dependent
reductions in physical condition and pregnancy rates, based
on resource acquisition during summer (Stewart et al.
2005). Those results indicate that density dependence is
driving population dynamics in this ecosystem, regardless
of elk spending the winter on our feedground rather than
traditional winter ranges (Stewart et al. 2005).
The northeast area consisted of 4 major plant communities: (1) mesic forest, (2) xeric forest, (3) xeric grassland,
and (4) logged forest (Stewart et al. 2002: Fig. 3). Plant
nomenclature follows Hitchcock and Cronquist (1996).
Mesic forests occur on north-facing slopes with overstory
composition dominated by grand fir (Abies grandis). Xeric
forests generally occur on south- and east-facing slopes;
tree composition consisted primarily of Ponderosa pine
(Pinus ponderosa), with the understory dominated by elk
sedge (Carex geyeri; Stewart et al. 2002). Xeric grasslands
occur primarily on south- and east-facing slopes; this plant
community was dominated by a few grasses such as Idaho
fescue (Festuca idahoensis), and bluebunch wheatgrass
(Agropyron spicatum), and forbs such as low gumweed
(Grindelia nana; Stewart et al. 2002). Logged-forest
communities composed areas where timber was harvested
during 1991–1992. Grand fir on Starkey suffered widespread mortality ([90%) from spruce budworm (Choristoneura occidentalis) during the late 1980s, and timber
was harvested in areas where most trees were killed
(Rowland et al. 1997; Stewart et al. 2002). Following
removal of trees, those areas were seeded with several
species of grasses including orchardgrass (Dactylis glomerata), and bluegrass (Poa spp.; Stewart et al. 2002).
Experimental design and statistical analyses
Fig. 1 The northeast study area with low density (west 4.1/km2) and
high density (east 20.1 elk/km2) of North American elk (Cervus
elaphus) on the Starkey Experimental Forest and Range, Oregon,
USA. Four major plant communities are indicated with locations of
herbivore exclosures
During 1997, we initiated an experiment to examine effects
of population density of elk on net aboveground plant
productivity and offtake of plant biomass by those large
herbivores. We created two populations of elk at high and
low density relative to K, in the northeast east and west
study areas on Starkey (Fig. 1). We assessed vegetation
123
306
responses to herbivory by examining elk populations
maintained at high population density near carrying
capacity (K) based on physical condition of elk on Starkey,
which we estimated to be 20.1 elk/km2; and at low densities near or below maximum sustained yield (MSY), which
we estimated at 4.1 elk/km2 (Stewart et al. 2006). The
high-density population was randomly assigned to the
northeast east study area and the low-density population to
the northeast west study area (designation of study areas is
provided in the description of the study areas). This
manipulation of population density of elk began in 1998,
which was a pretreatment year, and we stocked each study
area with moderate densities of elk (Stewart et al. 2006).
The experimental manipulation of high and low population
density began in 1999, but because of a mishap that year, a
gate was left open between study areas, elk densities were
moderately high (10.8 elk/km2) and low (6.6 elk/km2;
Stewart et al. 2006). Finally in 2000 and 2001, we
restricted access to our study areas and we maintained our
targeted high (20.1 elk/km2) and low (4.1 elk/km2) densities of elk (Stewart et al. 2006). Our analyses of species
composition of plants include 1999–2001, when exclosures
were present in all plant communities, except 6 grassland
sites. Thus, NAPP in grasslands during 1999 only was
restricted to areas with herbivory by elk, but grassland
exclosures were established for 2000 and 2001. Stewart
et al. (2006) provide a detailed rationale for this experimental design.
We placed exclosures (32 9 32 m) in mesic, logged,
and xeric forests, with 3 replications per community for
each high and low-density treatment. Exclosures placed in
xeric grasslands were 12 9 12 m to accommodate plant
communities that occurred in smaller patches (Fig. 1).
Stewart et al. (2006) used ten 1-m2 movable exclosures
located outside permanent exclosures in mesic, logged, and
xeric forests and five movable exclosures in xeric grasslands (McNaughton 1979; McNaughton et al. 1996). We
clipped 0.25-m2 quadrates inside and outside movable exclosures once a month during spring and summer to assess
productivity of vegetation (Stewart et al. 2006). We sampled permanent exclosures with 0.25-m2 quadrates at the
beginning and end of each season to examine NAPP in the
absence of herbivory and to compare with areas grazed by
elk (Stewart et al. 2006). We used those estimates of NAPP
in the presence and absence of herbivores for our comparisons with species diversity and richness of plants.
Because we were interested in the range of NAPP with
measures of species composition across the study area,
habitats were combined for those analyses. Our replications
consisted of permanent exclosure locations (inside and
outside), 3 exclosures in each of 4 plant communities per
treatment (high and low population density), for a total of
24 sampling sites. We understand that these sites are
123
Oecologia (2009) 161:303–312
replicated within our 2 density treatments (e.g., 12 sampling sites in each density treatment). Large herbivores,
including elk, do not use habitat uniformly, each of our
study areas is greater than 600 ha, and population density
and use of each of those sampling locations likely is
independent of the others.
We determined species composition of plants using
step-point transects inside and outside each permanent
exclosure (Bowyer and Bleich 1984; Bleich et al. 1997;
Stewart et al. 2006). We recorded a cover ‘‘hit’’ if the point
(\1 mm in diameter) fell within the canopy of a shrub or
on a stem or leaf of a plant. Each transect contained
approximately 200 step-points outside the exclosure and
100 step-points inside the exclosure, primarily because of
limited space within exclosures (Stewart et al. 2006).
Adequate sample size was determined by plotting the
number of species against cumulative number of points
sampled (Kershaw 1964; Geysel and Lyon 1980; Stewart
et al. 2006). We used the Shannon-Weiner formula (H0 ) to
estimate species diversity of plants from step-points for
cover inside and outside each exclosure (Ricklefs 1999;
Krebs 1999; Zar 1999).
Ungulates do not use habitat uniformly (Fretwell 1972;
Fryxell 1991; Kie et al. 2003) and some exclosure locations
in the high density area received less herbivory than some
locations in the low density area (Stewart et al. 2006).
Thus, the treatment density of elk for each study area was
inappropriate for estimating grazing intensity at each exclosure (Stewart et al. 2006). Accordingly, we used population densities of elk calculated outside each exclosure for
each year determined by Stewart et al. (2006). Those
densities were determined using radio telemetry locations
of elk, calculating density maps by using those locations,
which were then smoothed using kriging for each year
(Stewart et al. 2006). Eight adult females and four adult
males per year were equipped with radio collars in each
study area; error rate on telemetry locations was \50 m
(Stewart et al. 2002, 2006). Estimates of density at each
exclosure site were used to infer grazing intensity at each
sampling location outside each exclosure (Stewart et al.
2006). Stewart et al. (2006) provides detailed description of
methods used to estimate those densities. Mean densities at
exclosures in each of the study areas are provided in
Table 1.
We used linear regression to compare species richness
and diversity of plants with NAPP across our study areas
(Zar 1999). We also used linear regression to compare
plant species diversity with NAPP, both in the presence and
absence (exclosures only) of herbivory (Zar 1999). We fit
structural equation models (Grace 2006) for effects of elk
population density on plant species richness and diversity
(Fig. 4a). Structural equation models attempt to describe
the multivariate patterns between traits conditional on a
307
Table 1 Descriptive statistics (mean ± SD) for plant variables for
no herbivory (inside exclosures) by North American elk (Cervus
elaphus), and mean densities of herbivores in the low (4.1 elk/km2)
and high (20.1 elk/km2) density study areas
a
Variable
Species richness
Oecologia (2009) 161:303–312
Population density
(elk/km2)a
No herbivory Low
(n = 72)
herbivory
(n = 36)
0
5.0 ± 2.22
High
herbivory
(n = 36)
16.0 ± 6.32
NAPP (g m-2 day-1)a 0.60 ± 0.884 1.65 ± 1.855 1.27 ± 1.758
Plant species richness 17.0 ± 6.74 19.8 ± 6.88 20.0 ± 7.10
Plant species diversity 13.1 ± 5.43
(H0 )
13.3 ± 5.11
30
25
20
15
10
5
13.9 ± 5.03
Population densities of herbivores were determined at each sampling
site from animal locations to estimate levels of herbivory, on the
Starkey Experimental Forest and Range 1999–2001, Oregon, USA
0
b
From Stewart et al. (2006)
specific ordering of structural relationships that are inferred
from explicit biological hypotheses (Vile et al. 2006).
Thus, this method is appropriate for examining direct
effects of population density of elk on NAPP and indirect
effects of elk on species composition of plants (Vile et al.
2006; Bellingham and Sparrow 2009). We fit data to the
conceptual model as structural equation models using
AMOS student version 5.0.1. software (SPSS, Illinois,
USA).
-2
-1
0
1
2
-2
-1
0
1
2
3
4
5
6
4
5
6
30
25
Species diversity (H')
a
35
20
15
10
5
0
3
-2
-1
NAPP (g m day )
Results
We observed significant relationships between NAPP and
species richness (r2 = 0.042, CV = 36.91, F1,136 = 5.91,
P \ 0.001) and diversity (r2 = 0.054, CV = 38.07,
F1,136 = 7.76, P \ 0.001) across study areas. Those
regressions, however, had relatively low fit, based on low
r2 values, and low predictability based on low values for
coefficients of variation (Fig. 2). Population density of elk
did not affect species diversity directly (Table 1). We
observed no relationship between NAPP and species
diversity, in areas without herbivory (e.g., inside exclosures; F1,64 = 0.81, P = 0.373). Although in areas with
herbivory by elk (e.g., outside exclosures), the relationship
between NAPP and species diversity was highly significant
(r2 = 0.095, CV = 35.44, F1,70 = 7.33, P = 0.008;
Fig. 3); although again those regressions had relatively low
predictability and fit.
Structured equation models indicated significant quadratic relationships between population density and NAPP
(P = 0.007), and a positive relationship with diversity and
NAPP (P = 0.005; Fig. 4b). Structured equation models
performed well and indicated a significant, indirect relationship between population density and species richness
Fig. 2 Relationship between net aboveground primary production
(NAPP) with a plant species richness and b plant species diversity on
the Starkey Experimental Forest, La Grande, Oregon, USA, 1999–
2001. White circles indicate sampling locations in areas without
herbivory, gray circles are locations near exclosures in the study area
with low-density of elk, and black circles are locations near
exclosures in the study area with high-density of elk. Linear
regression indicated a positive relationship between NAPP and
species richness (a) (Y^ ¼ 17:51 þ 0:95x, r2 = 0.042, F1,136 = 5.91,
P = 0.016) and NAPP and species diversity (b) (Yˆ = 12.54 ? 0.81x,
r2 = 0.054, F1,136 = 7.76, P = 0.006)
and diversity of plants (Fig. 4b). R2 values for these models
were relatively low, indicating low predictability. We
observed significant paths between population density (and
population density2) with NAPP (P = 0.007), then NAPP
with species diversity (P = 0.005), which indicated a significant indirect effect of herbivory by elk on species
diversity of plants (Fig. 4b).
Discussion
Our research hypotheses were supported—in the areas
where NAPP was increased, we also observed increased
species richness and diversity of plants. Our data indicate
123
308
Oecologia (2009) 161:303–312
30
a
Population
density
Species diversity (H')
25
(Population
density)2
20
15
NAPP
10
eNAPP
5
0
-2
-1
0
1
2
3
-2
4
5
Plant species diversity
6
-1
NAPP (g m day )
Fig. 3 Relationship between net aboveground primary production
(NAPP) and plant species diversity in the presence of herbivory by
North American elk (outside exclosures only); linear regression also
indicated a positive relationship between NAPP and species diversity
of plants (Y^ ¼ 12:38 þ 0:86x, F1,70 = 7.33, r2 = 0.095, P = 0.008)
on the Starkey Experimental Forest, La Grande, OR, USA 1999–
2001. Gray circles are locations near exclosures in the study area with
low-density of elk, and black circles are locations near exclosures in
the study area with high-density of elk
ediversity
b
81.50
1166.06
5.26
(Population
density)2
Population
density
0.12
-0.01
NAPP
an indirect effect of population density of a large herbivore
on species composition through changing levels of NAPP;
low to moderate levels of herbivory had the greatest NAPP,
species richness, and species diversity in this montane
ecosystem. Regressions of NAPP and species diversity
inside exclosures without herbivory by elk were not significant, although the relationship between NAPP and
species diversity was significant in areas with herbivory by
elk. Ungulates do not use habitat uniformly and herbivory
outside exclosure locations varied greatly within density
treatments (Fretwell 1972; Fryxell 1991; Kie et al. 2003).
Indeed, in some instances, exclosure locations in the highdensity area received levels of herbivory similar to or
occasionally lower than locations in the low-density treatment. Thus increases in NAPP with low levels of herbivory
occurred at locations in both low and high density treatments (Figs. 2, 3). Simply comparing NAPP and species
richness or diversity by study area and treatment density
(e.g., no herbivory, high density, and low density) was too
coarse an examination to detect the fine-scale effects of
herbivory by elk on the plant community that we observed
in this study.
When we investigated indirect effects of population
density on species diversity using structured equation
models, our hypothesis was supported; we observed significant relationships of population density on NAPP and
through NAPP on species diversity. Results of those two
regressions and the structured equation model suggest that
herbivory may be driving the relationship between NAPP
123
2.11
0.82; R 2 = 0.05
0.81
Plant species diversity
12.54; R 2 = 0.05
25.59
Fig. 4 Structural equation model testing hypothesized relationships
for effects of population density of North American elk on net
aboveground primary production (NAPP) and species diversity of
plants. a Hypothetical structural equation model. One-headed arrows
represent causal relationships, and double-ended arrows indicate
correlation between variables (e.g., population density and population
density squared). Residual error variables (ei) represent effects of
unexplained causes (Vile et al. 2006). b Structural equation model
derived from model a. Path coefficients between variables are
unstandardized partial regression coefficients. Arrows not originating
from a variable represent residual error variances. Intercept values are
provided in italics next to boxes with variable names. Response boxes
contain the intercept and R2 value in italics. Arrows indicate
significant (P \ 0.005) paths between variables. Overall goodnessof-fit X22 ¼ 6:155; P = 0.05
and species diversity. In areas with herbivory by elk, the
relationship between NAPP and species diversity, although
significant, had relatively low predictability, most likely
because of great variation in NAPP across the differing
plant communities on the study area. We were unable to
determine, however, if increases in NAPP directly increase
diversity or if the effects of herbivores on interactions
Oecologia (2009) 161:303–312
among plants lead to both increases in NAPP and species
diversity. Further investigation to determine the mechanism driving those relationships among herbivores, NAPP,
and species diversity is warranted.
Stewart et al. (2006) suggested that where plants have a
long history of grazing and browsing, changes in species
composition with changing levels of herbivory may not be
evident. Indeed, Starkey has been grazed by both native
and domestic ruminants since the turn of the century
(Skovlin 1991). Since Starkey was fenced in 1987, cattle
were grazed in the northeast study area at moderate density
by researchers from Oregon State University. Although
cattle were removed from our study area prior to 1997
when we began our experiment, effects of long-term herbivory from domestic ruminants likely remain, such that
most of the species present in our study areas probably are
resilient to herbivory (Cingolani et al. 2005; Stewart et al.
2006). Communities that are more resilient to herbivory
may be more likely to respond to changes in densities of
herbivores indirectly through changes in NAPP rather than
rapidly changing species composition (Cingolani et al.
2005; Stewart et al. 2006). This process is likely to be
overlooked if the community is resilient to herbivory and
indirect responses are not examined.
We do not imply that large herbivores do not affect
species diversity directly; at high levels of herbivory, palatable species of plants have been shown to be negatively
affected in both eastern deciduous and western coniferous
forests, or those plants may be removed completely
(Rooney and Gross 2003; Rooney and Waller 2003; Beschta and Ripple 2008). Indeed, high densities of herbivores,
exhibiting density-dependent feedbacks on measures of
fitness (Gaillard et al. 2000), can drive successional changes leading to dominance of plant species of lower nutritional value, slower rates of nutrient cycling (Pastor and
Cohen 1997; Tremblay et al. 2005; Stewart et al. 2006),
and restricted regeneration of palatable shrubs and trees
such as black cottonwood (Populus trichocarpa) and bigleaf maple (Acer macrophyllum; Beschta and Ripple 2008).
Indeed, trophic cascades resulting from high densities of
herbivores following removal of carnivores has been
observed in multiple ecosystems (Terborgh et al. 2001;
Ripple and Beschta 2004; Beschta and Ripple 2008).
Mulder and Ruess (1998) demonstrated that the herbivory
by brant geese (Branta bernicla nigricans) altered the
competitive environment among plants in a subarctic salt
marsh community, and changed biomass allocation and
sexual reproduction in grazed plants. Moreover, high
population densities of geese (Anser sp.) resulting from
agricultural subsidies on winter ranges have resulted in
conversion of summer range in the Canadian arctic from
coastal salt marshes to mudflats (Jefferies 2000; Jefferies
and Rockwell 2002; Fox et al. 2005). At a local scale, those
309
geese have begun to show density-dependent feedbacks
resulting from lowered carrying capacity on the summer
range. Indeed, reduced growth and survival of young
indicate that those populations are rapidly approaching
carrying capacity (Person et al. 2003; Abraham et al. 2005;
Fox et al. 2005).
Ungulates act as keystone species in many ecosystems
that they inhabit, and at high population density those
populations have been observed to have negative effects on
nutrient cycling, plant productivity, and species composition (Augustine and McNaughton 1998; Rooney 2001;
Beschta and Ripple 2008), while simultaneously showing
negative feedbacks on physical condition and reproduction
associated with density dependence (Stewart et al. 2005).
Indeed, changing successional patterns resulting from high
levels of herbivory is not uncommon when herbivore
densities are close to carrying capacity. Tremblay et al.
(2005) suggested that because white-tailed deer (Odocoileus virginianus) switched from their primary forage balsam fir (Abies balsamea) to alternatives, such as lichens,
which incidentally are higher quality forage than fir
(Hodgman and Bowyer 1985; Ditchkoff and Servello 1998;
Jenks and Leslie 1989), those populations somehow
escaped density-dependent feedbacks. Nevertheless, those
authors also reported that the deer population appeared to
be stable over 25 years (Tremblay et al. 2005), which is a
strong indicator of density-dependent feedbacks regulating
the population near carrying capacity. Carrying capacity is
defined as the number of animals at or near equilibrium
with their food supply (Kie et al. 2003); however, that
definition does not mean that plant communities are static
or that primary forage types need to be in equilibrium with
the herbivore population. Herbivory can act as either a
stabilizing influence or an agent of disturbance depending
on where population density resides relative to carrying
capacity (Kie et al. 2003).
Species diversity appears to be a key factor in ecosystem
stability, and greater stability of plant community was
observed in ecosystems with high species diversity following perturbations (McNaughton 1977; Lepŝ et al. 1982;
Frank and McNaughton 1991; Tilman 1996; Mulder et al.
2001). Tilman (1996) suggested that plots with greater
species richness should have a higher probability of containing disturbance-resistance species, and that highdiversity mixtures should outperform low-diversity ones.
Mulder et al. (2001) reported no relationship between
species richness and biomass of bryophytes under constant
conditions; however, under drought conditions, biomass
increased strongly with increasing species richness. Thus,
Mulder et al. (2001) concluded that positive interactions
among plants may be an important mechanism linking high
diversity to high productivity under stressful environmental
conditions. This idea applies to herbivory as well, because
123
310
low-density populations of herbivores maintain high levels
of diversity in plant communities (McNaughton 1983,
1985; Tilman 1999; Collins et al. 1998; Collins and Smith
2006). Those more diverse communities likely would
contain mixtures of palatable and unpalatable species of
plants. Restoration of bison (Bison bison) to rangelands
decreased herbaceous cover and biomass, but increased
spatial heterogeneity and species richness as standing
biomass declined (Knapp et al. 1999; Jacobs and Naiman
2008).
Keystone species have the capacity to alter communities; whether those modifications are positive or negative
are driven by population densities of those species relative
to carrying capacity. The relationship between different
levels of herbivory and biodiversity appears to be strongly
linked to density-dependent processes of herbivores, and
resilience of the community to herbivory (Cingolani et al.
2005; Stewart et al. 2006). Low levels of herbivory,
resulting from herbivores at low population density relative
to carrying capacity, likely affect biodiversity in more
subtle ways than are typically addressed, particularly if
those communities are relatively resilient to herbivory.
Acknowledgments We appreciate the assistance of Starkey Project
Personnel, numerous volunteers, and technicians who helped with
most aspects of the study including sampling plants. We thank A. D.
Sparrow for advice and help with structured equation modeling. This
study was funded by the U.S. Forest Service, the Institute of Arctic
Biology at the University of Alaska Fairbanks, the Rob and Bessie
Welder Wildlife Foundation, Department of Natural Resources and
Environmental Science at University of Nevada Reno, the Rocky
Mountain Elk Foundation, and the Department of Biological Sciences
at Idaho State University. This is Rob and Bessie Welder Wildlife
Foundation contribution number 683. All aspects of this research
were approved by the Institutional Animal Care and Use Committee
at the University of Alaska Fairbanks (IACUC #01-34) and the
Starkey Project; and were in keeping with protocols adopted by the
American Society of Mammalogists for field research involving
mammals (Animal Care and Use Committee 1998).
References
Abraham KF, Jefferies RL, Alisauskas RT (2005) The dynamics of
landscape change and snow geese in mid-continent North
America. Glob Change Biol 11:841–855
Animal Care and Use Committee (1998) Guidelines for the capture,
handling, and care of mammals as approved by the American
Society of Mammalogists. J Mammal 79:1416–1431
Augustine DJ, McNaughton SJ (1998) Ungulate effects on the
functional species composition of plant communities: herbivore
selectivity and plant tolerance. J Wildl Manag 62:1165–1183
Bellingham PJ, Sparrow AD (2009) Multi-stemmed trees in montane
rain forests: their frequency and demography in relation to
elevation, soil nutrients, and disturbance. J Ecol (in press)
Berger J, Stacy PB, Bellis L, Johnson MP (2001) A mammalian
predator-prey imbalance: grizzly bear and wolf extinction affect
avian neotropical migrants. Ecol Appl 11:947–960
123
Oecologia (2009) 161:303–312
Beschta RL, Ripple WJ (2008) Wolves, trophic cascades, and rivers
in the Olympic National Park, USA. Ecohydrology 1:118–130
Bleich VC, Bowyer RT, Wehausen JD (1997) Sexual segregation in
Mountain sheep: resources or predation? Wildl Monogr 134:1–
50
Bowyer RT, Bleich VC (1984) Effects of cattle grazing on selected
habitats of southern mule deer. CA Fish Game 70:240–247
Bowyer RT, Person DK, Pierce BM (2005) Detecting top-down
versus bottom-up regulation of ungulates by large carnivores:
implications for conservation of biodiversity. In: Ray JC,
Redford KH, Steneck RS, Berger J (eds) Large carnivores and
the conservation of biodiversity. Island Press, Covelo, pp 342–
361
Caughley G, Krebs CJ (1983) Are big mammals simply little
mammals writ large? Oecologia 59:7–17
Cingolani AM, Noy-Meir I, Diaz S (2005) Grazing effects on
rangeland diversity: a synthesis of contemporary models. Ecol
Appl 15:757–773
Collins SL, Smith MD (2006) Scale-dependent interaction of fire and
grazing on community heterogeneity in tallgrass prairie. Ecology
87:2058–2067
Collins SL, Knapp AK, Briggs JM, Blair JM, Steinauer EM (1998)
Modulation of diversity by grazing and mowing in native
tallgrass prairie. Science 280:745–747
deCalesta DS, Stout SL (1997) Relative deer density and sustainability: a conceptual framework for integrating deer management
with ecosystem management. Wildl Soc Bull 25:252–258
Ditchkoff SS, Servello FA (1998) Litterfall: an overlooked food
source for wintering white-tailed deer. J Wildl Manage 62:250–
255
Fowler CW (1981) Density dependence as related to life history
strategy. Ecology 62:602–610
Fox AD, Madsen J, Boyd H, Kuijken, Norriss DW, Tombre IM,
Stroud DA (2005) Effects of agricultural change on abundance,
fitness components, and distribution of two arctic-nesting goose
populations. Glob Change Biol 11:881–893
Frank DA, McNaughton SJ (1991) Stability increases with diversity
in plant communities: empirical evidence from the 1988
Yellowstone drought. Oikos 62:360–362
Frank DA, McNaughton SJ (1993) Evidence for the promotion of
aboveground grassland production by native large herbivores in
Yellowstone National Park. Oecologia 96:157–161
Fretwell SD (1972) Populations in a seasonal environment. Princeton
University, Princeton
Fryxell JM (1991) Forage quality and aggregation by large herbivores. Am Nat 138:478–498
Gaillard JM, Festa-Bianchet M, Yoccoz NG, Loison A, Toigo C
(2000) Temporal variation in fitness components and population
dynamics of large herbivores. Annu Rev Ecol Syst 31:367–393
Geysel LW, Lyon LJ (1980) Habitat analysis and evaluation. In:
Schemnitz SD (ed) Wildlife techniques manual. Wildlife Society, Washington, DC, pp 305–327
Grace JB (2006) Structural equation modeling and natural systems.
Cambridge University Press, Cambridge
Hik DS, Jefferies RL (1990) Increases in the net above-ground
primary production of a salt-marsh forage grass: a test of the
predictions of the herbivore-optimization model. J Ecol 78:180–
195
Hirth DH (1977) Social behavior of white-tailed deer in relation to
habitat. Wildl Monogr 53:1–55
Hitchcock CL, Cronquist A (1996) Flora of the Pacific Northwest.
University of Washington, Seattle
Hodgman TP, Bowyer RT (1985) Winter use of arboreal lichens,
ascomycetes, by white-tailed deer, Odocoileus virginianus, in
Maine. Can Field Nat 99:313–316
Oecologia (2009) 161:303–312
Jacobs SM, Naiman RJ (2008) Large African herbivores decrease
herbaceous plant biomass while increasing plant species
richness in a semi-arid savanna toposequence. J Arid Env
72:889–901
Jefferies RL (2000) Allochthonous inputs: integrating population
changes and food-web dynamics. Trends Ecol Evol 15:19–22
Jefferies RL, Rockwell RF (2002) Foraging geese, vegetation loss,
and soil degradation in an arctic salt marsh. Appl Veg Sci 5:7–16
Jenks JA, Leslie DMJ (1989) Digesta retention of winter diets in
white-tailed deer Odocoileus virginianus fawns in Maine, USA.
Can J Zool 67:1500–1504
Kershaw KK (1964) Quantitative and dynamic ecology. Edward
Arnold, London
Kie JG, Bowyer RT, Stewart KM (2003) Ungulates in western forests:
habitat relationships, population dynamics, and ecosystem
processes. In: Zabel C, Anthony R (eds) Mammal community
dynamics in western coniferous forests: management and
conservation. Johns Hopkins University Press, Baltimore, pp
296–340
Knapp AK, Blair JM, Briggs JM, Collins SL, Harnett DC, Johnson
LC, Towne EG (1999) The keystone role of bison in North
American Tallgrass Prairie: bison increase habitat heterogeneity
and alter a broad array of plant, community, and ecosystem
processes. Bioscience 49:39–50
Krebs CJ (1999) Ecological methodology, 2nd edn. Addison Wesley
Longman, Menlo Park
Lepŝ J, Osbornová-Kosinová J, Rejmánek M (1982) Community
stability, complexity and species life history strategies. Vegetation 50:53–63
Mautz WW (1978) Sledding on a brushy hillside: the fat cycle in deer.
Wildl Soc Bull 6:88–90
McCullough DR (1979) The George Reserve deer herd: population
ecology of a K-selected species. University of Michigan, Ann
Arbor
McNaughton SJ (1977) Diversity and stability of ecological communities: a comment on the role of empiricism in ecology. Am Nat
111:515–525
McNaughton SJ (1979) Grazing as an optimization process: grassungulate relationships in the Serengeti. Am Nat 113:691–703
McNaughton SJ (1983) Serengeti grassland ecology: the role of
composite environmental factors and contingency in community
organization. Ecol Monogr 53:291–320
McNaughton SJ (1985) Ecology of a grazing ecosystem: the
Serengeti. Ecol Monogr 55:259–294
McNaughton SJ, Milchunas SJ, Frank DA (1996) How can net
primary productivity be measured in grazing ecosystems?
Ecology 77:974–977
McShea WJ (2005) Forest ecosystems without carnivores: when
ungulates rule the world. In: Ray JC, Redford KH, Steneck RS,
Berger J (eds) Large carnivores and the conservation of
biodiversity. Island Press, Washington, DC, pp 138–152
Molvar EM, Bowyer RT, Van Ballenberghe V (1993) Moose
herbivory, browse quality, and nutrient cycling in an Alaskan
treeline community. Oecologia 94:472–479
Mulder CPH, Ruess RW (1998) Effects of herbivory on arrowgrass:
interactions between geese, neighboring plants, and abiotic
factors. Ecol Monogr 68:275–293
Mulder CPH, Uliassi DD, Doak DF (2001) Physical stress and the
diversity-productivity relationships: the role of positive interactions. Proc Natl Acad Sci USA 98:6704–6708
Nicholson MC, Bowyer RT, Kie JG (2006) Forage selection by mule
deer: does niche breadth increase with population density? J
Zool Lond 269:39–49
Olff H, Ritchie ME (1998) Effects of herbivores on grassland plant
diversity. Trends Ecol Evol 13:261–265
311
Pastor J, Cohen Y (1997) Herbivores, the functional diversity of
plants species, and the cycling of nutrients in ecosystems. Theor
Pop Biol 51:165–179
Person DK, Bowyer RT, Van Ballenberghe V (2001) Density
dependence of ungulates and functional responses of wolves:
effects on predator-prey ratios. Alces 37:253–273
Person BT, Herzog MP, Ruess RW, Sedinger JS, Anthony RM,
Babcock CA (2003) Feedback dynamics of grazing lawns:
coupling vegetation change with animal growth. Oecologia
135:583–592
Persson IL, Danell K, Bergstrom R (2005) Different moose densities
and accompanied changes in tree morphology and browse
production. Ecol Appl 15:1296–1305
Ricklefs RE (1999) Ecology, 4th edn. Freeman, New York
Ripple WJ, Beschta RL (2004) Wolves, elk, willows, and trophic
cascades in the upper Gallatin Range of Southwestern Montana,
USA. For Ecol Manag 200:161–181
Rooney TP (2001) Deer impacts on forest ecosystems: a North
American perspective. Forestry 74:201–208
Rooney TP, Gross K (2003) A demographic study of deer browsing
impacts on Trillium grandiflorum. Plant Ecol 168:267–277
Rooney TP, Waller DM (2003) Direct and indirect effects of whitetailed deer in forest ecosystems. For Ecol Manag 181:165–176
Rowland MM, Bryant LD, Johnson BK, Noyes JH, Wisdom MJ,
Thomas JW (1997) The Starkey project: history, facilities, and
data collection methods for ungulate research. USDA For Ser
Gen Tech Rep PNW-GTR- 396:1–62
Ruess RW, McNaughton SJ (1987) Grazing and the dynamics of
nutrient and energy regulated microbial processes in the
Serengeti grasslands. Oikos 49:101–110
Ruess RW, Hik DH, Jefferies RL (1989) The role of lesser snow
geese as nitrogen processors in a sub-arctic salt marsh. Oecologia 79:23–29
Schoenecker KA, Singer FJ, Zeigenfuss LC, Binkley D, Menezes
RSC (2004) Effects of elk herbivory on vegetation and nitrogen
processes. J Wildl Manag 68:837–849
Simberloff D (1998) Flagships, umbrellas, and keystones: is single
species management passé in the landscape era? Biol Cons
83:247–257
Singer FJ, Schoenecker KA (2003) Do ungulates accelerate or
decelerate nitrogen cycling? For Ecol Manag 181:189–204
Skovlin JM (1991) Fifty years of research progress: a historical
document on the Starkey Experimental Forest and Range. USDA
For Ser Tech Rep PNW-GTR- 266:1–58
Stewart KM, Bowyer RT, Kie JG, Cimon NJ, Johnson BK (2002)
Temporospatial distributions of elk, mule deer, and cattle: do
stable isotopes reflect dietary niche?. J Mammal 83:229–244
Stewart KM, Bowyer RT, Kie JG, Dick BL, Ben-David M (2003)
Niche partitioning among mule deer, elk, and cattle: do stable
isotopes reflect dietary niche? Écoscience 10:297–302
Stewart KM, Bowyer RT, Dick BL, Johnson BK, Kie JG (2005)
Density dependence in North American elk: an experimental
test. Oecologia 143:85–93
Stewart KM, Bowyer RT, Ruess RW, Dick BL, Kie JG (2006)
Herbivore optimization in North American elk: consequences for
theory and management. Wildl Monogr 167:1–24
Terborgh J, Lopez L, Nunez P, Rao M, Shahabuddin G, Orhuela G,
Riveros M, Ascanio R, Adler GR, Lambert TD, Balbas L (2001)
Ecological meltdown in predator-free forest fragments. Science
294:1923–1926
Tilman D (1996) Biodiversity: population versus ecosystem sustainability. Ecology 77:350–363
Tilman D (1999) The ecological consequences of changes in
biodiversity: a search for general principles. Ecology 80:1455–
1474
123
312
Tremblay JP, Thibault I, Dussault C, Huot J, Cote SD (2005) Longterm decline in white-tailed deer browse supply: can lichens and
litterfall act as alternative food sources that preclude densitydependent feedbacks. Can J Zool 83:1087–1096
Vellend M (2004) Parallel effects of land-use history on species
diversity and genetic diversity of forest herbs. Ecology 85:3043–
3055
123
Oecologia (2009) 161:303–312
Vile D, Shipley B, Garnier A (2006) A structural equation model to
integrate changes in functional strategies during old-field
succession. Ecology 87:504–517
Zar JH (1999) Biostatistical analysis, 4th edn. Prentice-Hall, Upper
Saddle River